Conditional Expection on Sum of 2 Random Variables E(X|X+Y)=X+Y?

67 Views Asked by At

Suppose X is independent of Y, where X is any distribution, Y is Gaussian (0, $\sigma^2$). Do we have E(X|X+Y)=X+Y? Thank you!

1

There are 1 best solutions below

3
On

This is definetly not true in general. Note that $\mathbb{E}(X|X+Y)=X+Y$ implies that \begin{equation} \mathbb{E}(Y|X+Y)=0. \end{equation} For a counterexample, let $X$ and $Y$ be two independent Gaussian random variables. Then $\mathbb{E}(Y|X+Y)$ is again Gaussian (and not constantly equal to 0).